Agreement between rapid on-site evaluation and the final cytological diagnosis of salivary gland specimens.

OBJECTIVE: Rapid on-site evaluation (ROSE) has been shown to be beneficial in determining the adequacy and a preliminary diagnosis in multiple organ systems. However, little is known regarding the diagnostic values and agreement of ROSE compared with a final cytological diagnosis in salivary gland fine needle aspiration (FNA). The aim of the present study was to evaluate agreement and compare accuracy between ROSE and a final cytological diagnosis of salivary gland FNA, using a final histological diagnosis as a gold standard. METHODS: All patients with salivary gland lesions who underwent FNA with ROSE during 2009 to 2013 were evaluated. The patient demographic data, clinical characteristics, ROSE, final cytological diagnosis and final histological diagnosis were obtained. Agreement and accuracy were assessed by Kappa statistic and receiver-operating characteristic (ROC) curve analysis, respectively. RESULTS: Of 386 FNAs with ROSE, 248 (64%) lesions were from parotid glands, and 171 (44%) had histological follow-up results. Agreement between ROSE and the final cytological diagnosis was good to excellent (simple kappa = 0.76; linear weighted kappa = 0.81). There were discrepancies in 25 lesions (6.8%). Of these, 10 (2.7%) were major discrepancies, resulting in a change in clinical management. Only one of 36 malignant lesions by ROSE was downgraded to suspicious. No significant difference was found in accuracy between two interpretations (area under the curve 0.82 vs 0.84, P=.17). CONCLUSION: These findings suggest that, in addition to an excellent agreement and comparably good accuracy between ROSE and final cytological evaluation, malignant results of ROSE may be useful to facilitate an early clinical decision.

Atypical and suspicious categories in fine needle aspiration cytology of the breast: histological and mammographical correlation and clinical significance.

INTRODUCTION: This study aims to correlate fine-needle aspiration specimens diagnosed as C3 (atypical, probably benign) and C4 (suspicious, probably malignant) with histology and mammography, and to evaluate these two cytology categories in terms of diagnostic usefulness and patient management. METHODS: All fine-needle aspiration (FNA) specimens in categories C3 or C4 at the Maharaj Nakorn Chiang Mai Hospital, Thailand between 2000-2004 were reviewed. Results were correlated with available histological and mammographical studies. RESULTS: 148 FNA specimens were identified, comprising 43 category C3 and 105 category C4. Histology was available in 90 cases. 14 (64 percent) C3 cases showed benign histology on biopsy and eight (36 percent) were malignant. 13 (19 percent) C4 cases were benign on biopsy, whereas 55 (81 percent) were malignant. Mammographical studies were available in 56 of the histologically-proven cases. All seven cases with benign mammograms had benign histology, and all 26 cases called "highly suggestive of malignancy" were malignant on histology (five C3 and 21 C4). Of the 23 cases called "suspicious abnormality" on mammography, 14 turned out to be malignant on biopsy (one C3 and 13 C4). CONCLUSION: Our study supports maintaining cytology categories C3 and C4. About two-thirds of C3 cases were benign on biopsy whereas 81 percent of C4 cases were malignant (p-value is less than 0.001). There was complete correlation between histological and mammographical studies except those with equivocal mammograms. Our study supports the combined use of clinical, mammographical and cytological findings for optimal patient management. This is especially important for patients with C3 aspiration results, in order to avoid unnecessary surgery for benign lesions.

Different localization of dendritic cell reservoirs in human immunodeficiency virus-1 subtype B versus subtype E-infected lymph nodes.

The presence of p24 protein was studied in lymph nodes from human immunodeficiency virus (HIV)-positive patients affected by persistent generalized lymphadenopathy. Paraffin-embedded lymph node sections from 50 HIV-1 subtype E-infected lymph nodes from patients in Thailand and 25 HIV-1 presumably subtype B-infected lymph nodes from patients in the United States were immunostained with p24 HIV major core and capsid monoclonal antibodies using the streptavidin-biotin immunoperoxidase technique. Positivity for HIV p24 protein was detected in 20 of 22 HIV-1 subtype B infected nodes in which lymphoid follicles were present, with p24 staining demonstrating a reticular pattern within the germinal centers. Interestingly, no case from 50 clade E-infected lymph nodes containing lymphoid follicles had such a reticular pattern in the germinal centers. This difference could be explained by differential infection of subsets of dendritic cells by the two HIV-1 clades, or perhaps by different routes of initial HIV-1 transmission.

In vivo identification of Langerhans and related dendritic cells infected with HIV-1 subtype E in vaginal mucosa of asymptomatic patients.

In Thailand, the predominant HIV subtype is E, rather than Subtype B as in North America and Europe, and the predominant mode of transmission is heterosexual contact. Subtype E has the ability to replicate in vitro in Langerhans cells. We hypothesized that this cell type might constitute a reservoir for the HIV virus in vaginal mucosa of asymptomatic carriers. To examine this hypothesis, we compared vaginal tissue histology in HIV-1-seropositive cases with seronegative cases and determined the immunophenotype of HIV-1-infected cells, their numbers, and their distribution in vaginal mucosa. Vaginal biopsies were performed at four different sites from six asymptomatic HIV-1 Subtype E-infected persons and from six seronegative cases at necropsy and examined histologically. Immunophenotyping was performed using immunoperoxidase for Gag p24 HIV, CD3, CD20, CD68, CD1a, S-100 and p55 antigens and by double labeling, combining immunoperoxidase with alkaline phosphatase using pairs of the above antigens. Twenty of twenty-four vaginal biopsies (83.3%) from HIV-seropositive cases showed definite inflammation compared to five of twenty-four vaginal necropsies (20.8%) from HIV-seronegative cases. One third of HIV-seropositive biopsies (8/24) demonstrated p24-positive cells in the epithelium, whereas three-fourths (18/24) of the biopsies revealed p24-positive cells in the lamina propria. All seropositive patients showed positive cells in at least one biopsy, but not all biopsies contained positive cells. Infected cells were more frequently observed at sites of greater inflammation. The dendritic cell count in HIV-seropositive vaginal epithelium was significantly higher than that observed in the seronegative cases (P =.004). The majority of p24-positive cells in the vaginal epithelium were Langerhans cells (CD1a+/S-100+), whereas in the lamina propria, about half of p24-positive cells were Langerhans-related dendritic cells (p55+ and S-100+) and half were T lymphocytes. In conclusion, the increased propensity for heterosexual transmission of Subtype E may be related to vaginal inflammation, leading to the accumulation of Langerhans cells and related dendritic cells which, once infected with HIV, can act as a reservoir for further virus transmission.